The question as it stands is more about psychology than philosophy. Be that as it may, I would like to try to answer the question by starting with an assessment of the OP's claims that
... it is quite literally impossible to get an ought from an is. It is literally logically contradictory.
A contradiction being (on the propositional level) the assertion of a proposition A and its logical opposite NOTA (I used to use "~" on the fly since my keyboard doesn't have an easy function for putting down the rotated "L" instead; but anyway I've seen that "~" is supposed to mean something different from the rotated "L"), for an inference of OBA (as "it ought to be that A") from some non-OBB to be contradictory, the inference would have to be to both OBA and OB(NOTA), which I have never seen inferred as such.E
- I am obligated to do X; therefore, I ought to do X.
- There is an obligation to do X; therefore, I ought to do X.
- It is wrong to do X; therefore, I ought not do X.
... and so on, are all apparent cases where a sentence in is-form converts into a sentence in ought-form.∃ So as a historical community (of analytic philosophers), we moved away from just Hume's formulation of the issue, to Moore's, which doesn't at first have as much to do with metasyntactic problems with is-ought inferences. You could say that Hume's gloss of the issue was syntactic and so Moore's was semantic, then, and the semantic gloss is much more difficult to deal with, as long as you think no naturalistic/quasi-naturalistic definition of deontic semantics is sufficiently workable.
For example, in A Theory of Justice, John Rawls (sec. 40) speaks of categorical imperatives (and the two principles of justice as such imperatives) in terms of "expressions of our true selves' nature." Or then Kant said that, "Act according to the truth," is an analytic moral imperative (contrasted with the synthetic apriority of the categorical imperative proper). The gist of it is that a lot of people, going back ages and ages, define obligation in terms like, "I am obligated to do X if acting the opposite way would express a false belief, i.e. a false belief I would actually have." Or then, "Beliefs about what ought to be, are true, if and only if we ought to have those beliefs themselves." (Utilitarianism doesn't conform to this standard quite so strictly, since unknowable goodness is an option in that tradition.) So they'd say that there are "is"-facts that, coupled to the abstract "ought"-fact (the doxastic-deontic action-theoretic translation scheme), yield further "is"- and "ought"-facts.
Now then, if this is what someone means by ought (or obligation or whatever), then for such a person, a moral nihilist appears as someone whose relevant actions either fail to express true beliefs or which actively express false ones. So perhaps a moral nihilist appears to be a metaphysically dishonest person. Worse, if an axiological sentence like, "Object X has moral value," is not supposed to be true, even though not yet an "ought"-sentence, yet even so, the moral realist will look at the moral nihilist and think, "Oh, they won't murder me, perhaps, but it's not because they acknowledge my natural moral value; for them, not murdering me is as (un)well-motivated as not drinking tea, supposing that they don't like tea."
Granted, there is something superstitious about the worth of such fears. Even if everyone were an avowed moral realist de jure, it's not clear from our de facto more pluralistic historical vantage that no one would ever commit murder. From Socrates to Kant to Donald Davidson, a disbelief in unrestricted akrasia has been a recurrent mainstay in the analytic-minded thread of historical philosophy. This disbelief has even been prominent among Christian demonologists/hamartiologists (in the form of the theory that Satan did not sin by absolutely hating God in Itself but by some more convoluted default in "his" spiritual intellection). But so what strong difference does it make to our moral safety, if a few scattered people are born with cognitive deficits relating to moral concepts, and who therefore are "morally colorblind," so to say? Or even if it's just that they're mistaken, albeit culpably so; the point being, contrariwise, that we have generally as much reason to be afraid of being murdered by arrogant moral realists as by fickle moral nihilists (or various other adjectivized variants of either philosophical category).
EThough see the debate over agglomeration in deontic logic. Waiving agglomeration here, then, we would need to be able to infer OBA and NOT(OBA) to get a contradiction, here, and I am presently at a loss as to why inferring an ought from an is would result in inferring not-ought too.
∃Now when Kant inveighed against is-ought inference, he just meant that, "Such-and-such an action is being done (by a lot of people); therefore, it ought to be done," does not goes through. This is a reasonable point to make, but other is-ought inferences would follow a different template.